Abstract:Oil well productivity capacity is an important parameter in oilfield development, which is of great significance for efficient development. Traditional oil well productivity capacity prediction methods have a series of problems, such as limited application scope, large prediction errors, difficulty in characterizing changes under the influence of multiple factors. Aiming at these problems, a well productivity prediction method based on machine learning algorithm was proposed. Taking Bohai X oilfield as the research object, 12 factors affecting oil well productivity capacity were selected from three aspects: geology, engineering, and production. The degree of each factor influence on oil well productivity capacity was analyzed by using the mean decrease impurity (MDI) method, the feature parameters were sequentially excluded, and redundant features that do not affect the prediction accuracy of the model were removed. And then support vector machine (SVM) optimized by improved whale optimization algorithm (IWOA) was used to establish prediction model for oil well productivity capacity. The results show that the main control factors of oil well productivity capacity are: permeability, porosity, effective thickness, pressure draw-down, perforation thickness, fracturing sand addition amount, resistivity, oil saturation, sand addition strength and shale content. The model based on SVM optimized by the improved whale algorithm have an average error of 9.3%, while the model based on SVM optimized by grid search and whale algorithm have bigger errors, which are 21.7% and 15.7% respectively. Residual sum of squares (R 2 ) values for SVM optimized by grid search optimization, whale algorithm and improved whale algorithm are 0.372, 0.939 and 0.941 respectively. The model based on SVM optimized by the improved whale algorithm has higher accuracy in predicting oil well productivity capacity. Compared with existing literature, the MDI method was used to optimize the factors affecting oil well productivity, and IWOA was used to improve the accuracy of oil well productivity capacity prediction. The research results can provide reference for the well productivity capacity prediction.

Prediction of Oil Well Output Based on Polynomial Trend Surface Analysis

Predicting Initial Individual Well Production by Using Numerical Simulation

Tight Oil Well Productivity Prediction Model Based on Neural Network

Oil well productivity capacity prediction based on support vector machine optimized by improved whale algorithm

Application of an enhanced decision tree learning approach for prediction of petroleum production

Time-series well performance prediction based on Long Short-Term Memory (LSTM) neural network model

Oilfield analogy and productivity prediction based on machine learning: Field cases in PL oilfield, China

Productivity prediction method of shallow reservoir in Weibei Oilfield

Productivity Prediction Model of Tight Oil Reservoir Based on Particle Swarm Optimization–Back Propagation Neural Network

Production Capacity Prediction of Fractured Horizontal Wells Based on XGBoost

A Novel Model for Forecasting Production Performance in Waterflood Oil Reservoirs

Production prediction and main controlling factors in a highly heterogeneous sandstone reservoir: Analysis on the basis of machine learning

Application of Machine Learning for Productivity Prediction in Tight Gas Reservoirs

Using Nonlinear Regression Method in Oil Production Prediction

A Hybrid Oil Production Prediction Model Based on Artificial Intelligence Technology

The Research and Application of BP Network Tracking Model for Forecasting Oil Well Yield

Refined Prediction Method for Production Performance of Deep-Water Turbidite Sandstone Oilfield: A Case Study of AKPO in Niger Delta Basin, West Africa

Full-cycle prediction of gas production variations in gas wells using different production methods

A new prediction model on production performance for polymer flooding

A Model Based on the Random Forest Algorithm That Predicts the Total Oil–Water Two-Phase Flow Rate in Horizontal Shale Oil Wells

Production Prediction Model of Tight Gas Well Based on Neural Network Driven by Decline Curve and Data